The Agent Skills Directory

[SAFE]: The skill is advisory-only and explicitly limits its operation to read-only tools (Read, Grep, Glob), preventing unauthorized system modifications or data exfiltration.
[SAFE]: Instructions strictly forbid the exposure of API keys or management tokens, recommending the use of environment variables and server-side proxies for security.
[SAFE]: No obfuscation, remote code execution patterns, or suspicious dependencies were detected. All described workflows align with the stated purpose of managing CMS content releases.

cms-releases